首页 | 本学科首页   官方微博 | 高级检索  
     

定性系统单元分解仿真算法研究
引用本文:杨善林,梁昌勇. 定性系统单元分解仿真算法研究[J]. 合肥工业大学学报(自然科学版), 2000, 23(1): 21-27
作者姓名:杨善林  梁昌勇
作者单位:合肥工业大学,管理学院,安徽,合肥,230009;合肥工业大学,管理学院,安徽,合肥,230009
摘    要:标准的QSIM算法,由于基于全局状态表示来描述、推理系统的行为,对于复杂系统推理求解时会产生不可控分枝问题.文章提出一种系统分解的仿真算法,将定性模型分解成称之为单元的子模型.对单元分别应用QSIM推理,对单元间的关系通过约束传播使得子模型推理一致.这种系统分解仿真算法可大大降低QSIM算法的不可控分枝,提高系统的仿真效率.

关 键 词:QSIM算法  模型分解  定性推理
修稿时间:1999-09-24

The study of system decomposing simulation algorithm based on QSIM
YANG Shan-lin,LIANG Chang-yong. The study of system decomposing simulation algorithm based on QSIM[J]. Journal of Hefei University of Technology(Natural Science), 2000, 23(1): 21-27
Authors:YANG Shan-lin  LIANG Chang-yong
Abstract:Whenthe system is complex,and the standard QSIM algorithm is applied,intractable branchingproblems may occur, because the QSIM algorithm describes and handles system behaviorsbased on the global states. This paper proposes a system decomposing simulation algorithm.This algorithm divides the qualitative model into some sub-models called components,separately simulate the components with QSIM, and keeps the components consistent throughconstraint propagation. The algorithm may greatly decrease the intractable branches andimprove the simulation efficiency.
Keywords:QSIM algorithm  model decomposing  qualitative reasoning
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号